The present invention relates to certifying that a region on magnetic media is defective.
Data is written onto magnetic media, such as magnetic tape, for storage and subsequent retrieval. Generally, data is written onto the tape in contiguous blocks separated by gaps. The gaps delimit blocks and assist in finding the beginning of a particular block during read operations. Each block may be further divided into matrices, with each matrix representing a data segment. Occasionally, through a defect in manufacturing or wear, a defect in the magnetic media will prevent the successful writing of one or more matrices. Detecting and bypassing such defects is important to the efficient operation of a tape access system.
A typical tape access system includes a tape head with a write module and a read module trailing the write module in the direction of tape travel. The write module contains a plurality of write elements for simultaneously writing many data tracks. Similarly, the trailing read module contains a plurality of read elements for simultaneously reading the data tracks. Following each write element with a read element permits immediate read after write to verify that the data has been correctly written onto the tape. Write circuitry converts data into write module write signals. Read circuitry converts read module read signals into data. A drive controller controls one or more motors for moving the tape past the tape head in a particular direction and at a particular speed. A head position servo positions the head across the width of the tape to permit write and real elements access to appropriate data tracks. Head position relative to the tape is determined by reading servo tracks on the tape with servo read elements on the head.
If, during a write operation, data read by the trailing read head does not match the data just written by the write head, a write error has occurred. Write errors result from many causes including incorrect positioning of the tape head relative to the tape, incorrect tape velocity, defects on the tape, and the like. Typically, when a write error is detected, the tape is rewound and one or more attempts to rewrite the data are performed. After several unsuccessful attempts to write the data with proper tape velocity and tape head positioning, the problem is assumed to be a defect on the tape. In order to prevent incorrectly reading this region of the tape during a subsequent read operation and to prevent subsequent write operations to this region, the tape is marked or certified to indicate the defective area.
One method of defect certification is to record a fixed length tone pattern such as an erase gap over a short length of tape and then begin rewriting the data. This is repeated until the data block is successfully written. A second method is to continue write attempts until the far end of the tape defect is found. Special blocks are then written before and after the defect. These special blocks contain data permitting the tape system to compute the distance to the opposite end of the defective area.
If the write error resulted from head positioning difficulties, the write operation is immediately terminated to prevent overwriting adjacent data tracks. Hence, the first method of writing fixed length tone patterns is not appropriate. Instead, a technique similar to the second method used for tape defects is often employed. This may, however, result in partially written data blocks being left on the tape between the special marks.
Several problems exist with certifying defective regions on magnetic tape such as due to defects on the tape and servo track errors. First, the use of short erase blocks requires an excessive amount of time for long tape defects because the tape system attempts multiple rewrites after each short erase pattern. While the time may be decreased through using longer erase blocks, the resulting tape wasted by erase gaps covering non-defective tape decreases the density of data which can be stored on the tape. Second, partial data blocks recorded between special marks in the second method may be incorrectly interpreted as good data during a search for data blocks in a tape read operation. Third, both methods require time consuming rewind operations. What is needed is to certify magnetic media in a manner that does not require excessive time, does not waste non-defective tape, and does not leave partial data blocks in defective regions of the tape.
It is an object of the present invention to certify defective regions of magnetic media on the fly.
It is another object of the present invention to certify magnetic media without excessive tape rewind.
It is still another object of the present invention to certify magnetic media without leaving partial data blocks in defective regions of tape.
In carrying out the above objects and other objects and features of the present invention, a method for certifying that a track of magnetic media is defective is provided. The start of a defect is detected. A sequence of defect matrices is written over the defect with the write element in a read-after-write tape head. Each defect matrix may include a worst case data pattern. Each defect matrix in the sequence is read with a read element in the read-after-write head. A check is made to determine if an error exists in data read from each defect matrix. If an error is found, a check is made to determine that the defect extends through each defect matrix.
In an embodiment of the present invention, the number of consecutive defect matrices not containing an error is counted. The defect is determined to have ended when the number of consecutive error-free defect matrices exceeds a threshold.
In another embodiment of the present invention, a preset number of defect matrices is written following the last defect matrix containing the defect.
In still another embodiment of the present invention, the method includes counting the number of defect matrices prior to locating the defect end and terminating the defect certification if the number of counted defect matrices exceeds a threshold.
In yet another embodiment of the present invention, the method includes writing special characters preceding the defect indicating the start of the defect and writing special characters following the defect indicating the end of the defect.
In a further embodiment of the present invention, a preamble data pattern is written within the sequence of defect matrices to permit data clock resynchronization.
A system for certifying that a track of magnetic media is defective is also provided. The system includes a tape head for accessing the tape. The tape head includes a read module following a write module in a direction of tape travel past the tape head. The write module simultaneously writes at least one data track on the tape. The read module simultaneously reads at least one data track on the tape. The system also includes a write formatter for writing a sequence of defect matrices onto at least one data track. The number of defect matrices in the sequence is based on the defect length. The system also includes a read formatter which reads each defect matrix, detects an error in data read from the defect matrix, determines if the defect matrix contains a defect based on error detection, and determines the defect length based on defect matrices in the sequence determined to have a defect.
The above objects and other objects, features, and advantages of the present invention are readily apparent from the following detailed description of the best mode for carrying out the invention when taken in connection with the accompanying drawings.